N+M service backup mechanism based on IP switching

Abstract

The invention relates to an N+M service backup mechanism based on IP switching, characterized in that M backup machines adopt a multi-backup mechanism for N main machines, wherein the multi-backup mechanism provides service instead of the main machines in an IP switching way, the backup machines regularly detect the main machines, an ICMP (Internet Control Message Protocol) response request message is sent through a PING (Packet Internet Grope) demandcommand, a reply message for receiving an ICMP response is recorded, the messages can be used for judging faults of networks or the main machines, and carrying out IP switching and service replacement, thereby achieving a mechanism with M-backup effects, wherein N>=1 and M>=1. In addition, the invention adopts a scheme that the main machines have double IPs, which solves the problem of IP conflict generated after the main machines recover after the switching. A No.1 responsibility system is adopted in the aspects of dispatch and detection. The invention has the advantages of simple and flexible deployment, good expansibility, support of multi-backup, high stability, support of different services of N servers, good applicability, simple dispatch for multi-backup, good usability, less amount of mutual detection of servers, less network expense and the like. The invention has important significance for enhancing the stability of the traditional large system and has favorable effect in the processes of actual development, deployment, application, maintenance, and the like.

Description

technical field [0001] The invention relates to an N+M service backup mechanism based on IP switching, and belongs to the fault-tolerant field of service backup. Background technique [0002] At present, most of the backup and fault-tolerant schemes for systems with high reliability and stability requirements adopt such a scheme, that is, one host is configured with one backup, and both are required to install the same service program, and the switching conditions between the two are: It is detected through specific messages. If it is judged that the main machine fails, the standby machine will switch to the main machine to provide continuous services. Such a backup mechanism has such problems in many applications: [0003] (1) High cost. A host machine needs to be configured with a backup machine, which is twice the original hardware cost. In addition, for the computer room, a machine needs a place to be placed, which puts a lot of pressure on the space of the computer r...

AI Technical Summary

Problems solved by technology

[0004] (2) Backup services have limited protection

A standby machine can only back up one main machine, if the machine fails after the backup, the backup cannot be performed

[0005] (3) Only the same application service can be backed up

These messages are generally unique to these application systems and are not universal

[0007] (5) The scheduling mechanism is complex and requires a separate scheduling device

In the case of setting multiple backup machines for multiple devices, a separate scheduling device is generally used and contains a complex scheduling mechanism, so that the failure of the single scheduling device itself will cause the failure of the entire backup mechanism and increase the cost.

Images